Abstract:

In order to achieve rapid discrimination analysis of edible oil oxidation, common edible vegetable oils available in
the market were subjected to oxidation treatment and then divided by peroxide value and acid value according to the Chinese
national standard into two groups: unoxidized and oxidized for use as a reference set and a validation set. The qualitative
analysis of edible oils with different oxidation degrees was determined by electronic nose (E-nose) technology combined
with cluster analysis (CA), principal component analysis (PCA), and linear discriminant analysis (LDA). A model that
allowed rapid discrimination of edible oils with different oxidation degrees was built using CA, PCA, and LDA, respectively.
The recognition rates of all calibration models were 100%, and the recognition rates of CA, PCA, and LDA on validation set
were 95.8%, 98.9% and 100%, respectively. The E-nose technique was workable for identifying edible oil oxidation.

Key words: edible oils, oxidation, E-nose technique, discrimination analysis